Clostridium difficile infection (CDI) has been an increasingly common and complicated treatment issue for veteran and non-veteran patients over the past decade. One of the most difficult management issues is that of recurrent CDI. ~20% of patients who are successfully treated with either metronidazole or vancomycin will have a recurrent episode. Those who have already had a recurrent episode have ~50% chance of third diarrheal episode and some patients will have recurrent CDI for months or years. In addition, recurrent CDI has had a major impact on healthcare facilities because of frequent readmission rates to acute care facilities. Several host factors have been implicated as risks for recurrent CDI. However, there has been relatively little attention paid to pathogen or bacterial factors that might influence the risk of recurrence. The increase in rates of CDI and complications such as recurrent CDI over the past decade has coincided with the emergence of the epidemic C. difficile BI/027/NAP1 strain. Several potential virulence factors expressed by this strain have been proposed, but recent evidence suggests that binary toxin (CDT) produced by this strain and other clinically important (e.g., BK/078/NAP7) strains may play a critical adjunctive role in CDI pathogenesis. C. difficile binary toxin (CDT) is unrelated to the large, single unit glycosylating toxins, toxin A and B. This 2-component (binary) toxin acts by ADP-ribosylation of the cytoskeletal protein, actin, resulting in microtubular structural changes and epithelial cell membrane protrusions which increase adherence and colonization of C. difficile. Our research aims are designed to test the hypothesis that binary toxin increases the risk of recurrent CDI by two experimental lines of investigation. First, we will use the hamster model of CDI to demonstrate the effect of binary toxin by using isogenic mutants of C. difficile in which the binary gene toxin is knocked out. Second, we will use a newly developed immunoassay for binary toxin to test patient stool specimens and correlate the presence of fecal binary toxin with clinical recurrence of CDI. The specific aims of this proposal include: 1. Establish a reproducible hamster model of CDI recurrence. Aim 1a) Construct binary toxin mutants in the epidemic-associated BI and BK strains (CDTneg) for the baseline experiments in the hamster model Aim 1b) Establish a reproducible recurrence rate in the hamster model using binary toxin mutants (CDTneg) & confirm increased recurrence rates using respective parent (CDTpos) strains 2. Determine the mechanism of increased CDI recurrence due to binary toxin Aim 2a) Determine the mechanism in vitro using a human intestinal epithelial cell line Aim 2b) Determine the mechanism in vivo in hamsters 3. Correlate the expression of binary toxin in patients with CDI recurrence Aim 3a) Correlate fecal binary toxin detection with CDI recurrence Aim 3b) Correlate the presence of binary toxin-positive C. difficile strains with CDI recurrence